Kernel Density Estimation of Polygon Features

ChristopherBlinn — Sun, 20 Mar 2011 18:46:20 GMT

I am running into an issue trying to conduct a kernel density estimation on a polygon layer. Here is what I am doing:

I have linear features that have a spatial error in their location. To account for this I created a buffer polygon layer around each line. The buffer represents all possible areas the line feature could have existed. I want to run a kernel density estimation at multiple bandwidths (needing multiple outputs) on the polygon features. To do this, I converted the polygon layer to a raster (values of 1 for cells that are autocorrelated with the polygons, values of 0 for background). From there, I want to convert the raster back into either points or polylines, running a kernel density estimation requires either a point or polyline geometry for the input. I attempted each of these tools (raster to point and raster to polyline), and approximately 25% through each process, it terminates with the error 999999.

This was the only attempt I had thought of to calculate a kernel density estimation of polygon features. I am unaware of how to get passed the 999999 error, or how to create a more efficient process to calculate the kernel density estimation for polygons.

Thank you in advance for your time in helping me with this issue.

Re: Kernel Density Estimation of Polygon Features

ChristopherBlinn — Sun, 20 Mar 2011 19:21:13 GMT

I think I have found the problem.

When I created the raster from the polygons I made the cell sizes very small, to make the resolution as high as possible. Because of that, I have over 400 million cells, which means when I try to create a point file, I am exceeding the maximum size of a shapefile (about 70 million features). Going to reduce the resolution and run process again.

topic Re: Kernel Density Estimation of Polygon Features in ArcGIS Spatial Analyst Questions

Kernel Density Estimation of Polygon Features

Re: Kernel Density Estimation of Polygon Features